Short answer: when a wearable fan advertises "cooling", the number behind that claim is almost always the temperature of the metal plate — not the temperature of the air reaching the wearer. Those are different measurements, and the gap between them is large. On our bench, two cooled units differed by 3.3 °C on plate temperature after the same 30 seconds at maximum airflow, which is a useful comparison between two products and no guide at all to how much cooler either one makes a person feel.
What the measurement actually is
In our screening, one of three units had no thermoelectric cooling module at all; the other two did. Both cooled units were tested the same way — maximum airflow, 30 seconds of cooling, then the surface of the metal plate read with a K-type thermocouple taped to it.
That reading is a surface temperature. It is a real measurement and it is reproducible, which is why it is the one that ends up in marketing material: it is a single striking number, and it is cold. But it answers a question nobody asked. The wearer is not touching the plate. The wearer is standing in the airflow that has passed over the plate.
The part of the module that never appears in the claim
A thermoelectric module does not create cold. It moves heat from one face to the other. That has three consequences that matter for a device worn on a belt:
- The cold face is only one half of the device. Whatever is removed from the plate has to go somewhere, and in a wearable fan it goes into the hot face and then into the surrounding housing.
- The hot face sits against the wearer. This is the structural difference between a cooled wearable fan and a cooled appliance. A countertop device can dump its waste heat into a room. A belt-mounted device dumps it into the object touching the person who is already hot.
- The airflow is the only route the cold has to the wearer. Air passes over the plate, picks up what it can, and arrives at the skin. The plate being cold is a necessary condition, not a sufficient one.
This is why we treat "cooling" as a claim about a system rather than about a surface — and why the source of that claim should say which part of the system was measured.
What the two-unit comparison does tell you
Both cooled units reached a lower plate temperature than ambient, and unit 3's plate ended 3.3 °C lower than unit 2's after an identical 30 seconds — roughly 15% lower as a proportion of the higher reading. Held to the same method, that is a legitimate comparison: one module, or one thermal path, is doing more work per second than the other.
It is worth being explicit about the limit, because this is the step most published claims skip:
- No ambient temperature was recorded on the same record. Without it, a plate temperature cannot be turned into a temperature drop, and a temperature drop is the only version of the number that means anything to a wearer. 21.7 °C in a 24 °C room and 21.7 °C in a 32 °C room are entirely different products.
- No air temperature was recorded before and after the plate. That difference is the delivered effect, and it is always smaller than the plate-to-ambient difference, because moving air does not fully equilibrate with a surface it passes over.
- No hot-side temperature was recorded. A module that cools well and rejects badly is trading a cold plate for a warm housing.
None of that invalidates the comparison between the two units. It invalidates the jump from a plate temperature to a claim about how cool a person will feel.
What a cooling claim should come with
If a product is going to advertise cooling, four numbers make the claim checkable. Each one is a bench measurement rather than an opinion:
| Measurement | Why it is the one that matters |
|---|---|
| Air temperature entering the airflow, and again after it has passed the plate | This difference is the delivered effect — the only figure that corresponds to what the wearer feels |
| Ambient temperature and humidity during the test | Without them no temperature reading can be interpreted, including the plate's |
| Time to reach the recorded plate temperature, and how long it holds | A small module on a small battery reaches its coldest point briefly; a claimed instantaneous figure says nothing about the next twenty minutes |
| Hot-side or housing temperature rise over the same period | This is where the removed heat goes, and on a wearable device it goes next to the person |
A supplier who can produce all four is describing a cooling system. A supplier who can produce only the first column of the table above — a cold plate temperature, quoted alone — is describing a metal part.
Why this is a buying risk, not a curiosity
The failure mode here is not that the cooling does nothing. It is that the claim sets an expectation the airflow cannot meet, and the wearer is the one who discovers the difference:
- Expectation mismatch produces returns. A user who read "cooling" and expected cold air on a hot day has a complaint that no repair can resolve, because the product is working as designed.
- The comparison is unfair to good products. A unit with a modest module and an excellent airflow path can deliver more real relief than a unit with a colder plate and a weak duct — and would lose on the plate number alone.
- It hides the uncomfortable half. Two units can quote the same cold plate temperature while differing substantially in how much heat they pump into the housing against the wearer's body.
The fix is not to refuse cooling claims. It is to insist that a claim names what was measured, where, and for how long — which is the same standard we apply to every other number on the sheet.
Frequently asked questions
Is a cooler cold plate always better?
For the same product and the same airflow, yes — a lower plate temperature gives the air more to pick up. Across two different products it is not comparable on its own, because the duct, the airflow path and the hot-side rejection are doing at least as much work.
Why does a cold plate not feel as cold as its temperature suggests?
Because air passing over a cold surface does not fully take on that surface's temperature. The air arrives, exchanges heat for a fraction of a second, and leaves warmer than the plate and cooler than it started. The delivered difference is always smaller than the plate-to-ambient difference.
Where does the heat from a thermoelectric module go?
Into the hot face and from there into whatever it is attached to — in a wearable fan, the housing and the air around it. That heat is part of the product's behaviour and belongs in any honest cooling specification.
How do I check a cooling claim on a sample?
Two thermocouples: one on the plate, one in the airflow after the plate. Record the air-off and air-on readings at the same setting, note ambient temperature and humidity, and time how long the plate stays cold. Those four readings replace the single plate figure entirely.
Can a fan cool a person without a thermoelectric module?
Yes — airflow alone removes heat from skin by convection and evaporation. One of the three units we measured had no cooling module at all, and airflow of that kind is a real and measurable effect. It just should not be described as "cooling" alongside a plate temperature, because the two claims work differently.
If you are sourcing in this category, our temperature and airflow measurement requirements are written as testable figures inside our OEM/ODM programme. See also why air speed alone is only half an airflow measurement. Request a quotation with your target market, volume and required cooling performance.